Research on generic model control method for manipulator based on neural networks

The trajectory tracking control problem based on generic model control method for manipulator is addressed in this paper. Aiming at the problem that the relative degree of the input and output of manipulator system is not equal to 1, the sliding mode switching function is established according to the trajectory tracking error, and a method using sliding mode switching function as system virtual output is presented for generic model control law design, so that the trajectory tracking error moves to the sliding mode switching surface and eventually tends to zero. Since it is difficult to establish the precise dynamics model for the manipulator, the adaptive identification method is put forward based on RBF neural networks, which introduces sliding mode switching item to compensate for the approximation error of neural networks and ensure the system stability. For the sliding mode switching item is easy to cause control system chattering, the sliding mode switching gain adjustment method based on the exponential function is proposed. Finally, the effectiveness of the proposed methods is verified by simulation results.